Bipolar disorder (BD) is a serious psychiatric disorder affecting 1-3% of the population. Though family and twin studies support strong heritability, progress towards identifying specific genes has been slow. Our consortium has studied the genetics of bipolar disorder for over 20 years, and collected one of the largest samples of bipolar families and cases in the world. Though genomewide association studies (GWAS) have identified some genes and a significant polygenic component, these mapping approaches have had only limited success. It has been proposed that this missing heritability is transmitted in part through a large number of rare variants of strong genetic effect. Such rare variants might range in frequency from uncommon (<1%) to extremely rare (private mutations), and may include a variety of types including SNPs, indels and CNVs. GWAS, even with next-generation chips, would have limited power to detect such rare variation, and exome sequencing misses many structural variants. Fortunately, next generation sequencing technology has advanced at a staggering pace in the last few years, making whole genome sequencing a practical and affordable tool. We propose to make use of our large collection of families with bipolar disorder and a recent large linkage study conducted by our consortium, in combination with whole genome sequencing and targeted sequencing, to identify rare variants of strong effect that play a causative role in BD. Using our recent linkage study, we have identified 52 families with strong evidence for linkage and selected one member from each family most likely to harbor causative variants. In Aim 1, we will sequence the entire genome of these 52 subjects in order to identify rare variants of strong effect in the genomic regions in each case for which there is evidence of linkage; 12 regions in all. Variants of predicted strong functional effect will be sought in the bet linkage region for each family. In Aim 2, these candidate functional rare variants will be validated by Sanger sequencing and examined for cosegregation with disease in their family. In Aim 3, these variants will be used to select a prioritized list of candidate genes that will then undergo targeted sequencing in a sample of 1,500 BD cases and 1,500 controls. We hypothesize that additional rare variants of strong functional effect will be identified in those genes that convey genetic vulnerability to BD. We will use a sophisticated collapsed variant set analysis approach to test the hypothesis that these genes carry a greater mutational burden in the BD cases as compared to controls. All sequencing data will be made available to the scientific community and will provide an invaluable resource for future studies. The identification of vulnerability genes and mutations in BD will contribute to our understanding of its biological mechanism and facilitate the development of new treatments.